Adaptive Amoeboid Locomotion That Exploits Law of Conservation of Protoplasmic Mass

This paper intensively discusses how adaptive locomotion under unstructured and dynamically changing environment can be realized from the perspective of long-distance interaction and local interaction dynamics induced in bodies. To this end, we have focused on the most primitive yet flexible locomotion, amoeboid locomotion. Slime mold and amoeba are well known to exhibit remarkably adaptive behaviors, such as avoiding hazardous condition, and approaching nutrients and humidity, by dynamically changing their morphology. These behaviors induced so-called amoeboid movement, which is driven by the flexible epitheca (i.e. outer skin) and the protoplasm despite of the absence of a central nervous system or specialized organs. In light of these facts, we have conducted simulations of an amoeboid robot, particularly focusing on epitheca consisting of "real-time tunable springs" and "law of conservation of protoplasmic mass", the former of which is used to deal with "local interaction dynamics" and the latter of which is used to deal with "long-distance interaction". Simulation results indicate that the proposed model can induce highly adaptive locomotion according to the situation encountered by dynamically changing its morphology.